Sycl Academy LogoSycl Academy Logo
listShow Lessons List
  1. What is SYCL
  2. Enqueueing a Kernel
  3. Managing Data
  4. Handling Errors
  5. Device Discovery
  6. Data Parallelism
  7. Introduction to USM
  8. Using USM
  9. Asynchronous Execution
  10. Data and Dependencies
  11. In Order Queue
  12. Advanced Data Flow
  13. Multiple Devices
  14. ND Range Kernels
  15. Image Convolution
  16. Coalesced Global Memory
  17. Vectors
  18. Local Memory Tiling
  19. Further Optimisations
  20. Matrix Transpose
  21. More SYCL Features
  22. Functors
Lesson
Source
Solution
GitHub Iconedit

Data and Dependencies

In this exercise you will learn how to create a data dependency data flow graph in both the buffer/accessor and USM models.

1.) Define a data flow graph with the buffer/accessor model

Using everything you have learned in previous exercises create an application using the buffer/accessor model which has four kernel functions. These kernel functions can do any computation you like, but they should follow the following dependencies.

      (kernel A)
     /          \
(kernel B)  (kernel C)
     \          /
      (kernel D)

The important thing here is that kernels B and C must depend on kernel A, kernel D must depend on kernels B and C and kernels B and C can be executed in any order and even concurrently if the device permits. Note that in the buffer/accessor these dependencies are created implicitly using the creation of accessors.

Feel free to use any method of synchronization and copy back you like, but remember to handle errors.

2.) Define a data flow graph with the USM model

Now do the same again but using the USM model. Note that in the USM model dependencies are defined explicitly by chaining commands via events.

Again feel free to use any method of synchronization and copy back you like, but remember to handle errors.

Build And Execution Hints

For DevCloud via JupiterLab follow these instructions.

For DPC++: instructions.

For AdaptiveCpp: instructions.